- Email: [email protected]
A semicontinuous, high-performance liquid chromatography-based assay for stromelysin
ANALYTICALBIOCHEMISTRY
180,110-113
(1989)
A Semicontinuous, High-Performance Liquid Chromatography-Based Assay for Stromelysin Richard
Harrison,
Department
Jennifer
of Enzymology,
Teahan,
and Ross Stein
Merck Sharp and Dohme Research Laboratories,
P.O. Box 2000, Rahway,
New Jersey 07065
ReceivedOctober 24,1988
A search for low molecular weight peptide substrates for the metalloendoproteinase, human fibroblast stromelysin, resulted in the discovery that substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-LeuMet-NHP) is a substrate for this enzyme and is cleaved exclusively at the Gin’-Phe’ bond. On the basis of this observation, a semicontinuous HPLC-based assay was developed that monitors the production of the hydrolysis product, fragment 7-l 1 (SP’-“). Steady-state velocities for the production of SP’-l’ have been determined as a function of substrate concentration and obey simple, Michaelis-Menten kinetics. For a l-ml reaction volume, V,,, = (2.4 nmol SP’-“/min)/pg protein and K,,, = 0.38
mM.
0 1989 Academic
Press,
Inc.
Stromelysin is a metalloendoproteinase that is synthesized and secreted by mammalian fibroblasts (l-4). Its ability to degrade proteoglycan and other proteins of the cartilage matrix forms the basis for the view that stromelysin plays a pathogenic role in osteo- and rheumatoid arthritis (5,6). Inhibitors of this enzyme may therefore be of therapeutic value in the treatment of these diseases. As the initial part of a long-term program to study stromelysin catalysis and inhibition, we needed to develop an easily automated assay based on the hydrolysis of a low molecular weight peptide substrate. Although low molecular weight substrates for stromelysin have been reported (2,7), none of these studies contain kinetic analyses. To efficiently screen a large number of peptides for their ability to serve as stromelysin substrates and then to determine kinetic parameters for the reactions of those substrates that were identified, we have employed high-performance liquid chromatography. Although the use of HPLC for the study of enzyme kinetics has increased in recent years (8,9), the assays which have been reported all involve aliquot withdrawal with off-line quenching prior to chromatographic analysis and there110
fore result in labor-intensive and time-consuming assays. To avoid these shortcomings, we used an HPLC autosampler and on-line quenching to provide semicontinuous analysis of enzyme reaction solutions. Using this methodology, we identified substance P (SP,l Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-LeuMet-NHz) as a substrate for stromelysin and determined that it is cleaved at a single site, the Gin’-Phe7 bond. This discovery forms the basis of the assay that we describe herein. MATERIALS
AND METHODS
Chemicals. Tris-HCl, sodium azide, and Brij 35 were obtained from Sigma (St. Louis, MO). Peptides were purchased from Sigma and Bachem and used without further purification. Calcium chloride was purchased from Mallinkrodt (Paris, KY). House-distilled water was deionized using a Milli-Q water purification system (Millipore, Millford, MA). HPLC grade acetonitrile (CH&N) and trifluoroacetic acid (TFA) were obtained from Fisher Scientific (Pittsburgh, PA). Substrate. SP and SP fragment 7-11 (SP7-11) were obtained from Sigma and used without further purification. Solutions of SP were made in a pH 7.5 buffer containing 20 mM Tris-HCl, 10 mM CaClz, 0.02% NaN3, and 0.05% Brij 35. Enzyme. Stromelysin was supplied by Dr. M. Lark of the Department of Biochemical and Molecular Pathology, Merck and Co. Dr. Lark and co-workers purified the enzyme as the zymogen from the culture media of interleukin l-stimulated human gingival fibroblasts (10). This enzyme was identified as stromelysin (l-4) by two criteria: (i) agreement of N-terminal sequence analysis of the zymogen and trypsin-activated form with the literature sequence (2,lO); and, (ii) Western immu1 Abbreviations used: SP, substance P; TFA, trifluoroacetic APMA, p-aminophenylmercuric acetate; SDS-PAGE, sodium cyl sulfate-polyacrylamide gel electrophoresis.
acid; dode-
0003.2697/89 $3.00 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
CHROMATOGRAPHIC
-
ASSAY
1752 1 2.109
r
-
-
-
1.590 '2.207
-
FOR
STROMELYSIN
111
Kinetic assays. In a typical kinetic measurement, 750 ~1 of a buffered solution of substrate in a 1.5-ml glass autosampler vial was thermally equilibrated to 25°C for 15 min in the thermostated compartment of the autosampler. Two hundred and fifty microliters of a 25°C solution of activated enzyme was then added to this vial to give a final stromelysin concentration of 2 pg/ml. The reaction solution was mixed by inversion and placed back in the autosampler. Immediately after the initiation of the reaction, and at 3-min intervals thereafter, 20-~1 aliquots were withdrawn by the autosampler and injected onto the column.
= -
-
3.804
-
4.457
FIG. 1. Chromatography of SP after a 60-min incubation with either buffer (upper panel) or 2 fig/ml stromelysin (lower panel). Assay conditions are as described under Materials and Methods. Peak identity (retention time): buffer salts and SP’-” (1.7 min); Brij 35 (2.1 min); NaN, and SP (2.2 min); and, SP”’ (3.8 min).
noblotting with a monoclonal antibody raised against recombinant stromelysin (10). Prostromelysin was supplied to us as a 10 pg/ml solution in a pH 7.5 buffer containing 20 mM Tris-HCl, 300 mM NaCl, 10 mM CaCl,, 0.02% NaN3, and 0.05% Brij 35. The zymogen was activated by one of two methods. In the first, the zymogen was incubated at 37°C for 4 h with 2 mM p-aminophenylmercuric acetate (APMA). After activation, the solution was stored at 4°C until use. Alternately, the zymogen was incubated at 37°C with 12 nM trypsin for 1 h. Inactivation of the trypsin after stromelysin activation was accomplished by the addition of a 20-fold molar excess of soybean trypsin inhibitor bound to agarose (Sigma, St. Louis, MO). Both methods produce an active stromelysin species that appears homogeneous on SDS-PAGE with a molecular weight of 48,000 (10). Chromatographic conditions. The chromatographic apparatus consisted of a Waters 510 isocratic pump (Waters Association, Millford, MA), an Hitachi Model 655A-40 autosampler fitted with a temperature control option (E. M. Science, Cherry Hill, NJ). Temperature was maintained at 25.0 f O.l”C by a Lauda RM6 refrigerated circulating water bath (Brinkmann, Westbury, NY). Absorbance was measured at 215 nm using a Kratos Spectroflow 783 variable wavelength detector (ABI Analytical, Ramsey, NJ). Peak integrations were performed on a Vista 401 chromatographic data station (Varian Association, Walnut Creek, CA). The chromatographic column was a Whatman RAC II Partisil 5 C8 column (10 cm X 4.6 mm i.d.). Mobile phase was an aqueous solution of 0.1% TFA (pH 2) and CH&N at a ratio of 67/33 (v/v). The flow rate was 1.0 ml/min.
RESULTS
To identify a substrate for stromelysin, we tested a number of commercially available peptides for their susceptibility to stromelysin-catalyzed hydrolysis. Of the peptides that were examined as potential substrates, only SP was hydrolyzed. All the other peptides were hydrolyzed to an extent less than 5% in 20 h at a peptide concentration of 1 mM and an enzyme concentration of 2 pg/ml. Included among these were the peptides that were reported by other investigators to be stromelysin
FIG. 2. Chromatogram showing the sequential injection of a reaction solution of 3 mM SP and 2 fig/ml stromelysin. Conditions are as outlined under Materials and Methods. The highlighted peaks correspond to the hydrolysis product, SP7-“.
112
HARRISON,
TEAHAN,
AND
STEIN
.
-c
0 0
20
40
60
SO
100
Tk4E WWTES)
substrates (2,7): dynorphin l-7 (Tyr-Gly-Gly-PheLeu-Arg-Arg), dynorphin 1-8 (Tyr-Gly-Gly-PheLeu-Arg-Arg-Ile), and N-(2,4-dinitrophenyl)-Pro-LeuGly-Ile-Ala-Gly-Arg-NH,. The hydrolysis of SP by stromelysin is illustrated in the chromatograms of Fig. 1. This figure demonstrates that SP is rapidly hydrolyzed by stromelysin, but is stable toward hydrolysis in neutral, aqueous solution. The cleavage site within SP was identified by two methods. First, using three chromatographic systems, we observed co-chromatography of one of the hydrolysis products
i
2
3
4
5
6
[Sj (ml4
FIG. 4. Substrate concentration ties for the stromelysin-catalyzed ume of 1.0 ml, the final protein was varied from 0.2 to 5.0 mM. independent kinetic experiments.
400
600
800
1000
TIME (MINUTES)
FIG. 3. Reaction progress curve for the production of SP7-” during the reaction of 3 mM SP and 2 pg/ml stromelysin. Concentrations of SP7-” were generated from integration of chromatogram peaks and a calibration curve (see text for details). The total amount of SP hydrolyzed to products in this experiment was 6%. The solid line represents the velocity of the steady-state reaction that follows the pre-steadystate transient phase.
00
200
dependence of steady-state velocihydrolysis of SP. In a reaction volconcentration was 2 pg/ml and SP The figure contains data from four
FIG. 5. Reaction of 55 FM SP with 2 fig/ml stromelysin. Solid circles correspond to the disappearance of SP, while open circles correspond to the generation of hydrolysis product, SP7-“. Best-fit lines were calculated by nonlinear least-squares fit of the data to a first-order rate law.
with authentic SP7-11. Second, we collected the eluate volumes corresponding to the two hydrolysis peaks and found, by amino acid analysis, that they corresponded to SPlm6 and SP7-“. Together, these results indicate that SP is cleaved at a single site and that this site is the Gin’-Phe7 bond. Additional support for a single site of cleavage is provided below. Figure 2 is a chromatogram generated in a kinetic experiment conducted as described under Materials and Methods. The time-dependent increase in the hydrolysis The areas under these product, SP7-11, is highlighted. peaks were converted to molar concentrations using a linear calibration curve generated from concentrations of SP7-11that ranged from 3 to 600 PM (data not shown). From this hydrolysis data, a reaction progress curve of [SP7-‘l] vs time was constructed and appears in Fig. 3. Reaction progress curves for the stromelysin-catalyzed hydrolysis of SP all display the pre-steady-state “burst” phase that is evident in Fig. 3. While the mechanistic origin of the burst is presently unclear, trivial explanations have been eliminated. We believe that the reaction of stromelysin and SP may represent an example of substrate-induced enzyme hysteresis (11). Despite the kinetic complexity of these reactions, we were able to estimate steady-state velocities as functions of substrate concentration. Results of four independent kinetic experiments appear in Fig. 4. From a nonlinear leastsquares fit of this data to the Michaelis-Menten equation, V,,, and K,,, can be estimated as (2.4 + 0.7 nmol SP7-11/min)/pg protein and 380 + 90 PM, respectively.2 ’ We believe that the scatter around the best-fit the large standard deviations determined for the we report result from two experimental difficulties.
line in Fig. 4 and kinetic parameters First, it is difficult
CHROMATOGRAPHIC
ASSAY
In another experiment, shown in Fig. 5, we monitored both SPi-ll accumulation and SP disappearance at an initial substrate concentration of 55 FM and an enzyme concentration of 2 Kg/ml. Since in this experiment [S] < K,,, , these progress curves are pseudo first order in substrate, with the observed rate constant equalling V,,,/ K,. From the curve corresponding to product accumulation, an observed rate constant was calculated by nonlinear least-squares analysis to be (3.67 + 0.07) X 10s3 min-‘. For substrate disappearance, the observed rate constant is (3.34 k 0.06) X 10m3 mini’. The near identity of these values not only supports a single site of cleavage but also validates the HPLC-based methodology reported herein. DISCUSSION
In this report, we demonstrate that SP is a substrate for human fibroblast stromelysin and that hydrolysis of SP occurs at a single site, the Gin’-Phe7 bond. Publications from other laboratories suggest similarities between stromelysin and thermolysin and report cleavage by stromelysin at Gly-Leu bonds (2,7). We find no evidence of cleavage of SP at the Phe7-Phe’ bond, as would be anticipated for an enzyme having a substrate specificity similar to that of thermolysin, nor do we detect hydrolysis of the Glyg-Leul’ bond. Also, we report the first kinetic analysis of a stromelysin-catalyzed reaction. Our steady-state analysis indicates that the reaction of stromelysin with SP follows Michaelis-Menten kinetics over a substrate concentration range of 0.1-5.6 mM. Our ability to identify SP as a substrate for stromelysin, and to then determine its kinetics of hydrolysis, was based on HPLC technology that has only recently been applied to enzyme kinetic problems. Earlier limitations,
to estimate steady-state velocities from the nonlinear progress curves that result during the stromelysin-catalyzed hydrolysis of SP (see Fig. 3). This is especially so at low substrate concentrations, where this problem is confounded by the steady-state requirement that only a small portion (ideally less than 5%) of substrate be converted to prod. uct. Second, we find that treatment of the zymogen of stromelysin with APMA or trypsin does not uniformly produce the same amount of active enzyme.
FOR
113
STROMELYSIN
such as inadequate temperature control of autosamplers, labor intensive analyses, and the inability to precisely quantitate results, have slowed the development of appropriate methodology for accurate enzyme kinetic investigations. In the present study, we took advantage of a number of important developments to overcome these limitations. First, the problem of temperature control was solved by using a temperature-regulated HPLC autosampler to house our reaction solutions. Second, the labor-intensive quenching step, that is integral to most HPLC-based enzyme assays, is avoided here by injecting aliquots of the reaction solution directly into the column eluant. The low pH and high organic content of the mobile phase effectively terminates stromelysin-catalyzed reactions. Third, the concentrations of product and substrate we use to generate reaction progress curves are precisely quantitated by automated peak integration coupled with a calibrated conversion of areas to concentrations. REFERENCES 1. Murphy, G., Nagase, and Brinckerhoff, C. E. (1988) Collagen Related Res. 8,389-391. 2. Murphy, G., Gavrilovic, J., and McAlpine, C. (1986) Proteinases in Inflammation and Tumor Invasion, Gruyter, New York. 3. Okada, Y., Nagase, H., and Harris, E. D. (1986) Trans. Assoc. Amer. Physicians 99,143-153. 4. Okada, Y., Harris, 731-741.
E. D., and Nagase,
H. (1988)
5. Ehrlich, M., Armstrong, A., Treadwell, J. Rheunatol. 14,30-33. 6. Cawston, Arthritis
T., Mercer, E., deSilva, Rheum. 27,285-290.
7. Shaw, A., Roberts, R., and Wolanin, tives in Anti-Inflammatory Therapies 67-80, Raven Press, New York. 8. Halfpenny, 9. Rossomondo,
M.,
Biochem.
J. 254,
B., and Mankin,
H. (1987)
and
B. (1984)
Hazelman,
D. J. (1988) in New Perspec(Lewis, A., Ed.), Vol. 12, pp.
A. (1984) Chromatogr. Sci. 28,285-302. E. (1987) LC-GC 5,880-889.
10. Lark, M. W., VanMiddlesworth, J., Walakovits, L., and Azzolina, T. (1988) Proceedings of the Fourth International Conference of the Inflammation Research Association, Abstract No. 145, McNeil Pharmaceuticals. 11. Neet, N. K., and Ainslie, G. R. (1980) in Methods in Enzymology (Purich, D. L., Ed.), Vol. 64, pp. 192-226, Academic Press, San Diego, CA.
180,110-113
(1989)
A Semicontinuous, High-Performance Liquid Chromatography-Based Assay for Stromelysin Richard
Harrison,
Department
Jennifer
of Enzymology,
Teahan,
and Ross Stein
Merck Sharp and Dohme Research Laboratories,
P.O. Box 2000, Rahway,
New Jersey 07065
ReceivedOctober 24,1988
A search for low molecular weight peptide substrates for the metalloendoproteinase, human fibroblast stromelysin, resulted in the discovery that substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-LeuMet-NHP) is a substrate for this enzyme and is cleaved exclusively at the Gin’-Phe’ bond. On the basis of this observation, a semicontinuous HPLC-based assay was developed that monitors the production of the hydrolysis product, fragment 7-l 1 (SP’-“). Steady-state velocities for the production of SP’-l’ have been determined as a function of substrate concentration and obey simple, Michaelis-Menten kinetics. For a l-ml reaction volume, V,,, = (2.4 nmol SP’-“/min)/pg protein and K,,, = 0.38
mM.
0 1989 Academic
Press,
Inc.
Stromelysin is a metalloendoproteinase that is synthesized and secreted by mammalian fibroblasts (l-4). Its ability to degrade proteoglycan and other proteins of the cartilage matrix forms the basis for the view that stromelysin plays a pathogenic role in osteo- and rheumatoid arthritis (5,6). Inhibitors of this enzyme may therefore be of therapeutic value in the treatment of these diseases. As the initial part of a long-term program to study stromelysin catalysis and inhibition, we needed to develop an easily automated assay based on the hydrolysis of a low molecular weight peptide substrate. Although low molecular weight substrates for stromelysin have been reported (2,7), none of these studies contain kinetic analyses. To efficiently screen a large number of peptides for their ability to serve as stromelysin substrates and then to determine kinetic parameters for the reactions of those substrates that were identified, we have employed high-performance liquid chromatography. Although the use of HPLC for the study of enzyme kinetics has increased in recent years (8,9), the assays which have been reported all involve aliquot withdrawal with off-line quenching prior to chromatographic analysis and there110
fore result in labor-intensive and time-consuming assays. To avoid these shortcomings, we used an HPLC autosampler and on-line quenching to provide semicontinuous analysis of enzyme reaction solutions. Using this methodology, we identified substance P (SP,l Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-LeuMet-NHz) as a substrate for stromelysin and determined that it is cleaved at a single site, the Gin’-Phe7 bond. This discovery forms the basis of the assay that we describe herein. MATERIALS
AND METHODS
Chemicals. Tris-HCl, sodium azide, and Brij 35 were obtained from Sigma (St. Louis, MO). Peptides were purchased from Sigma and Bachem and used without further purification. Calcium chloride was purchased from Mallinkrodt (Paris, KY). House-distilled water was deionized using a Milli-Q water purification system (Millipore, Millford, MA). HPLC grade acetonitrile (CH&N) and trifluoroacetic acid (TFA) were obtained from Fisher Scientific (Pittsburgh, PA). Substrate. SP and SP fragment 7-11 (SP7-11) were obtained from Sigma and used without further purification. Solutions of SP were made in a pH 7.5 buffer containing 20 mM Tris-HCl, 10 mM CaClz, 0.02% NaN3, and 0.05% Brij 35. Enzyme. Stromelysin was supplied by Dr. M. Lark of the Department of Biochemical and Molecular Pathology, Merck and Co. Dr. Lark and co-workers purified the enzyme as the zymogen from the culture media of interleukin l-stimulated human gingival fibroblasts (10). This enzyme was identified as stromelysin (l-4) by two criteria: (i) agreement of N-terminal sequence analysis of the zymogen and trypsin-activated form with the literature sequence (2,lO); and, (ii) Western immu1 Abbreviations used: SP, substance P; TFA, trifluoroacetic APMA, p-aminophenylmercuric acetate; SDS-PAGE, sodium cyl sulfate-polyacrylamide gel electrophoresis.
acid; dode-
0003.2697/89 $3.00 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
CHROMATOGRAPHIC
-
ASSAY
1752 1 2.109
r
-
-
-
1.590 '2.207
-
FOR
STROMELYSIN
111
Kinetic assays. In a typical kinetic measurement, 750 ~1 of a buffered solution of substrate in a 1.5-ml glass autosampler vial was thermally equilibrated to 25°C for 15 min in the thermostated compartment of the autosampler. Two hundred and fifty microliters of a 25°C solution of activated enzyme was then added to this vial to give a final stromelysin concentration of 2 pg/ml. The reaction solution was mixed by inversion and placed back in the autosampler. Immediately after the initiation of the reaction, and at 3-min intervals thereafter, 20-~1 aliquots were withdrawn by the autosampler and injected onto the column.
= -
-
3.804
-
4.457
FIG. 1. Chromatography of SP after a 60-min incubation with either buffer (upper panel) or 2 fig/ml stromelysin (lower panel). Assay conditions are as described under Materials and Methods. Peak identity (retention time): buffer salts and SP’-” (1.7 min); Brij 35 (2.1 min); NaN, and SP (2.2 min); and, SP”’ (3.8 min).
noblotting with a monoclonal antibody raised against recombinant stromelysin (10). Prostromelysin was supplied to us as a 10 pg/ml solution in a pH 7.5 buffer containing 20 mM Tris-HCl, 300 mM NaCl, 10 mM CaCl,, 0.02% NaN3, and 0.05% Brij 35. The zymogen was activated by one of two methods. In the first, the zymogen was incubated at 37°C for 4 h with 2 mM p-aminophenylmercuric acetate (APMA). After activation, the solution was stored at 4°C until use. Alternately, the zymogen was incubated at 37°C with 12 nM trypsin for 1 h. Inactivation of the trypsin after stromelysin activation was accomplished by the addition of a 20-fold molar excess of soybean trypsin inhibitor bound to agarose (Sigma, St. Louis, MO). Both methods produce an active stromelysin species that appears homogeneous on SDS-PAGE with a molecular weight of 48,000 (10). Chromatographic conditions. The chromatographic apparatus consisted of a Waters 510 isocratic pump (Waters Association, Millford, MA), an Hitachi Model 655A-40 autosampler fitted with a temperature control option (E. M. Science, Cherry Hill, NJ). Temperature was maintained at 25.0 f O.l”C by a Lauda RM6 refrigerated circulating water bath (Brinkmann, Westbury, NY). Absorbance was measured at 215 nm using a Kratos Spectroflow 783 variable wavelength detector (ABI Analytical, Ramsey, NJ). Peak integrations were performed on a Vista 401 chromatographic data station (Varian Association, Walnut Creek, CA). The chromatographic column was a Whatman RAC II Partisil 5 C8 column (10 cm X 4.6 mm i.d.). Mobile phase was an aqueous solution of 0.1% TFA (pH 2) and CH&N at a ratio of 67/33 (v/v). The flow rate was 1.0 ml/min.
RESULTS
To identify a substrate for stromelysin, we tested a number of commercially available peptides for their susceptibility to stromelysin-catalyzed hydrolysis. Of the peptides that were examined as potential substrates, only SP was hydrolyzed. All the other peptides were hydrolyzed to an extent less than 5% in 20 h at a peptide concentration of 1 mM and an enzyme concentration of 2 pg/ml. Included among these were the peptides that were reported by other investigators to be stromelysin
FIG. 2. Chromatogram showing the sequential injection of a reaction solution of 3 mM SP and 2 fig/ml stromelysin. Conditions are as outlined under Materials and Methods. The highlighted peaks correspond to the hydrolysis product, SP7-“.
112
HARRISON,
TEAHAN,
AND
STEIN
.
-c
0 0
20
40
60
SO
100
Tk4E WWTES)
substrates (2,7): dynorphin l-7 (Tyr-Gly-Gly-PheLeu-Arg-Arg), dynorphin 1-8 (Tyr-Gly-Gly-PheLeu-Arg-Arg-Ile), and N-(2,4-dinitrophenyl)-Pro-LeuGly-Ile-Ala-Gly-Arg-NH,. The hydrolysis of SP by stromelysin is illustrated in the chromatograms of Fig. 1. This figure demonstrates that SP is rapidly hydrolyzed by stromelysin, but is stable toward hydrolysis in neutral, aqueous solution. The cleavage site within SP was identified by two methods. First, using three chromatographic systems, we observed co-chromatography of one of the hydrolysis products
i
2
3
4
5
6
[Sj (ml4
FIG. 4. Substrate concentration ties for the stromelysin-catalyzed ume of 1.0 ml, the final protein was varied from 0.2 to 5.0 mM. independent kinetic experiments.
400
600
800
1000
TIME (MINUTES)
FIG. 3. Reaction progress curve for the production of SP7-” during the reaction of 3 mM SP and 2 pg/ml stromelysin. Concentrations of SP7-” were generated from integration of chromatogram peaks and a calibration curve (see text for details). The total amount of SP hydrolyzed to products in this experiment was 6%. The solid line represents the velocity of the steady-state reaction that follows the pre-steadystate transient phase.
00
200
dependence of steady-state velocihydrolysis of SP. In a reaction volconcentration was 2 pg/ml and SP The figure contains data from four
FIG. 5. Reaction of 55 FM SP with 2 fig/ml stromelysin. Solid circles correspond to the disappearance of SP, while open circles correspond to the generation of hydrolysis product, SP7-“. Best-fit lines were calculated by nonlinear least-squares fit of the data to a first-order rate law.
with authentic SP7-11. Second, we collected the eluate volumes corresponding to the two hydrolysis peaks and found, by amino acid analysis, that they corresponded to SPlm6 and SP7-“. Together, these results indicate that SP is cleaved at a single site and that this site is the Gin’-Phe7 bond. Additional support for a single site of cleavage is provided below. Figure 2 is a chromatogram generated in a kinetic experiment conducted as described under Materials and Methods. The time-dependent increase in the hydrolysis The areas under these product, SP7-11, is highlighted. peaks were converted to molar concentrations using a linear calibration curve generated from concentrations of SP7-11that ranged from 3 to 600 PM (data not shown). From this hydrolysis data, a reaction progress curve of [SP7-‘l] vs time was constructed and appears in Fig. 3. Reaction progress curves for the stromelysin-catalyzed hydrolysis of SP all display the pre-steady-state “burst” phase that is evident in Fig. 3. While the mechanistic origin of the burst is presently unclear, trivial explanations have been eliminated. We believe that the reaction of stromelysin and SP may represent an example of substrate-induced enzyme hysteresis (11). Despite the kinetic complexity of these reactions, we were able to estimate steady-state velocities as functions of substrate concentration. Results of four independent kinetic experiments appear in Fig. 4. From a nonlinear leastsquares fit of this data to the Michaelis-Menten equation, V,,, and K,,, can be estimated as (2.4 + 0.7 nmol SP7-11/min)/pg protein and 380 + 90 PM, respectively.2 ’ We believe that the scatter around the best-fit the large standard deviations determined for the we report result from two experimental difficulties.
line in Fig. 4 and kinetic parameters First, it is difficult
CHROMATOGRAPHIC
ASSAY
In another experiment, shown in Fig. 5, we monitored both SPi-ll accumulation and SP disappearance at an initial substrate concentration of 55 FM and an enzyme concentration of 2 Kg/ml. Since in this experiment [S] < K,,, , these progress curves are pseudo first order in substrate, with the observed rate constant equalling V,,,/ K,. From the curve corresponding to product accumulation, an observed rate constant was calculated by nonlinear least-squares analysis to be (3.67 + 0.07) X 10s3 min-‘. For substrate disappearance, the observed rate constant is (3.34 k 0.06) X 10m3 mini’. The near identity of these values not only supports a single site of cleavage but also validates the HPLC-based methodology reported herein. DISCUSSION
In this report, we demonstrate that SP is a substrate for human fibroblast stromelysin and that hydrolysis of SP occurs at a single site, the Gin’-Phe7 bond. Publications from other laboratories suggest similarities between stromelysin and thermolysin and report cleavage by stromelysin at Gly-Leu bonds (2,7). We find no evidence of cleavage of SP at the Phe7-Phe’ bond, as would be anticipated for an enzyme having a substrate specificity similar to that of thermolysin, nor do we detect hydrolysis of the Glyg-Leul’ bond. Also, we report the first kinetic analysis of a stromelysin-catalyzed reaction. Our steady-state analysis indicates that the reaction of stromelysin with SP follows Michaelis-Menten kinetics over a substrate concentration range of 0.1-5.6 mM. Our ability to identify SP as a substrate for stromelysin, and to then determine its kinetics of hydrolysis, was based on HPLC technology that has only recently been applied to enzyme kinetic problems. Earlier limitations,
to estimate steady-state velocities from the nonlinear progress curves that result during the stromelysin-catalyzed hydrolysis of SP (see Fig. 3). This is especially so at low substrate concentrations, where this problem is confounded by the steady-state requirement that only a small portion (ideally less than 5%) of substrate be converted to prod. uct. Second, we find that treatment of the zymogen of stromelysin with APMA or trypsin does not uniformly produce the same amount of active enzyme.
FOR
113
STROMELYSIN
such as inadequate temperature control of autosamplers, labor intensive analyses, and the inability to precisely quantitate results, have slowed the development of appropriate methodology for accurate enzyme kinetic investigations. In the present study, we took advantage of a number of important developments to overcome these limitations. First, the problem of temperature control was solved by using a temperature-regulated HPLC autosampler to house our reaction solutions. Second, the labor-intensive quenching step, that is integral to most HPLC-based enzyme assays, is avoided here by injecting aliquots of the reaction solution directly into the column eluant. The low pH and high organic content of the mobile phase effectively terminates stromelysin-catalyzed reactions. Third, the concentrations of product and substrate we use to generate reaction progress curves are precisely quantitated by automated peak integration coupled with a calibrated conversion of areas to concentrations. REFERENCES 1. Murphy, G., Nagase, and Brinckerhoff, C. E. (1988) Collagen Related Res. 8,389-391. 2. Murphy, G., Gavrilovic, J., and McAlpine, C. (1986) Proteinases in Inflammation and Tumor Invasion, Gruyter, New York. 3. Okada, Y., Nagase, H., and Harris, E. D. (1986) Trans. Assoc. Amer. Physicians 99,143-153. 4. Okada, Y., Harris, 731-741.
E. D., and Nagase,
H. (1988)
5. Ehrlich, M., Armstrong, A., Treadwell, J. Rheunatol. 14,30-33. 6. Cawston, Arthritis
T., Mercer, E., deSilva, Rheum. 27,285-290.
7. Shaw, A., Roberts, R., and Wolanin, tives in Anti-Inflammatory Therapies 67-80, Raven Press, New York. 8. Halfpenny, 9. Rossomondo,
M.,
Biochem.
J. 254,
B., and Mankin,
H. (1987)
and
B. (1984)
Hazelman,
D. J. (1988) in New Perspec(Lewis, A., Ed.), Vol. 12, pp.
A. (1984) Chromatogr. Sci. 28,285-302. E. (1987) LC-GC 5,880-889.
10. Lark, M. W., VanMiddlesworth, J., Walakovits, L., and Azzolina, T. (1988) Proceedings of the Fourth International Conference of the Inflammation Research Association, Abstract No. 145, McNeil Pharmaceuticals. 11. Neet, N. K., and Ainslie, G. R. (1980) in Methods in Enzymology (Purich, D. L., Ed.), Vol. 64, pp. 192-226, Academic Press, San Diego, CA.